Center for Expanded Data Annotation and Retrieval (CEDAR) - Overall

描述(申请人提供)：大数据革命要求生物医学科学家能够定位、分析和集成目前遍布生物医学的大数据集。只有当实验数据集在网上可用，并用元数据注释它们时，才有可能进行这种工作。元数据解释了数据是如何组织的，数据代表了什么，以及数据是如何收集的。扩展数据注释和检索中心(CEDAR)将利用社区驱动的元数据标准最近的增长，开发创新的计算方法，以简化元数据注释的创作和使用。我们的具体目标集中于与研究人员社区合作，以标准化通过生物医学研究生成的数据的描述；创建用于描述实验室研究的元数据模板的开发、评估、使用和改进的计算集体；开发全面和开放的元数据储存库，该储存库将为将推动我们中心的大部分技术的学习算法提供信息；培训生物医学界使用元数据和Cedar的资源；以及在ImmPort的背景下评估我们的工作，ImmPort是NIAID支持的多化验数据储存库，将提供端到端的机会来演示和验证我们的想法。我们预计，从人类免疫学项目联盟开始，然后是BD2K中心联盟，然后是斯坦福数字存储库，用户社区将不断增长，直到我们发展了广泛的用户基础，导致用于注释在线数据集的元数据的质量发生了可衡量的变化。我们项目的总体描述概述了雪松的活动和总体具体目标。

DESCRIPTION (provided by applicant): The Big Data revolution requires that biomedical scientists be able to locate, analyze, and integrate the large datasets that now pervade biomedicine. Such work is possible only when experimental datasets are made available online and when they are annotated with metadata that explain how the data are organized, what the data represent, and how the data were collected. The Center for Expanded Data Annotation and Retrieval (CEDAR) will take advantage of the recent growth in community-driven metadata standards to develop innovative computational methods to ease the authoring and use of metadata annotations. Our specific aims focus on working with communities of investigators to standardize descriptions of the data generated through biomedical studies; creating a computational collective for development, evaluation, use, and refinement of metadata templates for describing laboratory studies; developing a comprehensive and open repository of metadata that will inform the learning algorithms that will drive much of our Center's technology; training the biomedical community in the use of metadata and in CEDAR's resources; and evaluating our work in the context of ImmPort, an NIAID-supported multi-assay data repository that will offer end-to-end opportunities to demonstrate and validate our ideas. We anticipate a growing community of users, starting with the Human Immunology Project Consortium, then the BD2K Center Consortium, then the Stanford Digital Repository, growing until we have developed a wide user base leading to measurable changes in the quality of the metadata used to annotate online datasets. The Overall description of our project provides a synopsis of CEDAR's activities and overall specific aims.